The 90-Second Rule: How Long Emotions Actually Last

9 min read · Reviewed against current psychological research

Someone tells you the science says an emotion only lasts ninety seconds, and that anything past that is a choice you’re making. It arrives as a kindness and lands as an accusation. You’ve been upset about this for three days. Apparently that’s on you.

The claim is everywhere: therapy blogs, LinkedIn posts, corporate wellbeing decks, coaching newsletters. It sounds like neuroscience. It uses a number, which makes it sound measured.

It isn’t a finding. This article covers where it came from, what researchers found when they measured how long emotions last, and the part of the rule that survives contact with the evidence.

Where the number comes from

The source is Jill Bolte Taylor’s 2006 book My Stroke of Insight. Taylor is a neuroanatomist who had a severe haemorrhagic stroke in 1996 and wrote a memoir about the experience and her recovery.

In it she describes emotions as neurochemical events that flush from the bloodstream in around ninety seconds, with anything after that sustained by the thoughts you keep having about the trigger.

Two things about that are worth holding at once. Taylor is a credentialled scientist and the book is a serious, valuable account of neurological injury and recovery. And a memoir is not a study. The ninety-second figure is an observation about her own subjective experience, made during and after a catastrophic brain injury, which is an unusual state to generalise from.

No study has tested it. There is no paper you can look up, no sample, no measurement of when an emotion ends. The number entered circulation with the authority of neuroscience attached and none of the apparatus.

What happens when you measure it

Researchers have asked how long emotions actually last, and the answer is not one number.

Verduyn and Lavrijsen had participants recall recent emotional episodes and report how long each lasted, across 27 different emotions. Sadness was the longest, averaging around 120 hours. Shame was the shortest, at roughly half an hour. Surprise, fear, disgust, boredom, irritation and relief clustered at the brief end (Verduyn & Lavrijsen, 2015).

Sit with the size of that gap. The longest emotion lasted something like 240 times as long as the shortest. A single figure covering all of them is not an approximation, it’s a category error.

One honest caveat about that study, since this article is about representing evidence properly. It relied on people recalling and estimating durations rather than tracking them in real time, and recall of that kind is imperfect. What it establishes firmly is the enormous variation between emotions. Treat the specific hours as indicative rather than exact.

The finding holds up with better methods

That recall limitation is worth taking seriously, so it matters that the pattern also appears in studies that didn’t rely on memory.

An earlier pair of experience sampling studies had participants report on fear, anger, joy, gratitude and sadness episodes daily, as they happened, rather than recalling them afterwards. Across all five emotions, two things predicted a longer episode: how important the eliciting situation was, and how intense the emotion was at onset (Verduyn et al., 2009).

Different method, different emotions, same conclusion about importance. That convergence is what makes the finding trustworthy in a way a single study wouldn’t be.

Notice also what neither study found: a common duration. If emotions cleared on a fixed chemical timer, duration would be roughly constant and these predictors would have nothing to explain.

What actually determines how long it lasts

The more useful part of that research is why the durations differ so much.

Two factors accounted for almost half the variability between emotions. The first is how important the triggering event was. The second is how much the person ruminated about it (Verduyn & Lavrijsen, 2015).

Which explains the sadness result rather neatly. Sadness tends to follow events that matter, and events that matter are the ones you turn over repeatedly. It isn’t that sadness has a slower chemical half-life. It’s that sadness attaches to things worth thinking about.

Rumination as an amplifier is well documented in its own right. It exacerbates depression, sharpens negative thinking, impairs problem solving, and erodes social support (Nolen-Hoeksema et al., 2008). It’s not a neutral way of processing. It extends and deepens what it touches.

The part of the rule that holds up

Here’s why this myth has been so durable, and why the debunk shouldn’t go further than the evidence allows.

The rule says: the initial surge passes, and what keeps a feeling alive afterwards is the thinking you do about it. Strip out the ninety seconds and the neurochemistry, and that second half is very close to what the research found. Rumination is one of the two things that predicts duration.

So the rule is wrong about the mechanism and directionally right about the lever. That’s an unusual combination and it’s the honest verdict. People who found it useful were not fooled about everything.

What it gets wrong matters though, because the specific error causes harm.

The damage the number does

It turns duration into a verdict. If the science says ninety seconds, then still being upset on Thursday means you chose it. That’s a short step from being told your grief is self-indulgent, and it’s exactly the wrong message to give someone whose sadness is doing the ordinary work of processing something that mattered.

It conflates rumination with feeling. Thinking about an important event is not automatically rumination. Reflection that moves somewhere, makes meaning, or reaches a decision is a different process from the repetitive loop that goes nowhere. The rule treats all post-90-second thought as a failure of will.

It sets an unreachable benchmark. If you believe emotions should be done in a minute and a half, every ordinary human timeline reads as evidence you’re broken. That’s a reliable way to add shame to sadness, and shame is not a helpful ingredient.

What to do with a feeling that’s outstaying its welcome

The research points at something more useful than a stopwatch.

Check which process you’re in. Ask whether the thinking is going anywhere. Reflection tends to produce new information, a shift, or a decision. Rumination repeats. If you’ve had the same thought four times with no movement, that’s the loop rather than the processing.

Expect long emotions after important events. If something mattered, a long tail is the normal response rather than a malfunction. Our guide to emotional regulation covers what to do while you wait it out.

Interrupt the loop rather than the feeling. Emotion regulation research distinguishes strategies by where they intervene, from changing the situation through to changing how you think about it, and they don’t work equally well (Gross, 2015). The target here isn’t the sadness, it’s the repetition. Changing what you’re doing, moving your body, or talking to someone tends to break a loop more reliably than deciding to stop feeling something.

Name it precisely. Shame and sadness behave very differently and get confused constantly. Knowing which one you’re in tells you what to expect. Our guide to emotional self-awareness covers why precise naming changes the experience.

When to talk to a professional

Emotion duration is not a diagnostic criterion and nothing here is a threshold you should measure yourself against.

Worth raising with someone qualified: if low mood has persisted most of the day, most days, for two weeks or more; if rumination has become the default rather than an occasional visitor; or if the feeling isn’t attached to an event you can identify. Rumination in particular responds to specific approaches, which is a better reason to seek help than any number of hours.

The takeaway

Emotions don’t run on a ninety-second timer. Across 27 of them, the range ran from about half an hour to about five days, and what separated the short from the long was how much the event mattered and how much it got replayed.

That’s a better model anyway, because it points at something you can influence. The stopwatch version offers only a verdict on how long you’ve taken. The research version says: this lasted because it mattered, and it’s lasting longer because you keep returning to it, and the second half is workable.

If you’ve been carrying something for days, you’re not failing at a ninety-second process. You’re having a normal response to something that counted.

Frequently asked questions

Sources

1. Verduyn, P., & Lavrijsen, S. (2015). Which emotions last longest and why: the role of event importance and rumination. Motivation and Emotion, 39(1), 119–127.
2. Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3(5), 400–424.
3. Taylor, J. B. (2006). My Stroke of Insight: A Brain Scientist’s Personal Journey. Viking. (Origin of the 90-second claim; a memoir, not a study.)
4. Gross, J. J. (2015). Emotion regulation: current status and future prospects. Psychological Inquiry, 26(1), 1–26.
5. Verduyn, P., Delvaux, E., Van Coillie, H., Tuerlinckx, F., & Van Mechelen, I. (2009). Predicting the duration of emotional experience: two experience sampling studies. Emotion, 9(1), 83–91.